中国激光, 2024, 51 (8): 0811001, 网络出版: 2024-03-29  

基于多元定标法的矿物铀元素光纤式激光诱导击穿光谱定量分析【增强内容出版】

Quantitative Measurement of Uranium in Ore Using Fiber‑Optic Laser‑Induced Breakdown Spectroscopy and Multivariate Calibration
作者单位
1 西安交通大学电气绝缘与电力设备国家重点实验室,陕西 西安 710049
2 中国原子能科学研究院,北京 102413
3 西安交通大学电子物理与器件教育部重点实验室,陕西 西安 710049
摘要
针对核工业辐照环境下铀等元素组分远程测量的需求,基于光纤式激光诱导击穿光谱系统,研究了铀矿石样品中铀元素谱线强度的影响因素及其定量测量方法。在实验中,采用光纤传输30 mJ脉冲激光并将其聚焦在铀矿石粉末压片上产生激光诱导等离子体,然后通过同轴光纤回传等离子体自发光。首先,研究了气氛环境对铀元素谱线的影响。结果显示:氦气气氛下铀质量分数为0.425%的样本的U II 409.013 nm谱线的信噪比相比空气气氛下提高了1.37倍,同时,氦气气氛下铀质量分数为0.0726%的样本的U II 409.013 nm谱线的信噪比达到了8.9。其次,根据谱线信噪比、信背比和净强度将系统的探测延时优化至1000 ns。最后,提出了一种基于内标法原理的多元线性定标方法,通过引入基质元素与铀的谱线比值进行回归,使得铀的检测限达到142 mg/kg,定量限达到426 mg/kg。与基于谱线峰值的单变量回归方法相比,所提多元线性定标方法将铀元素的定标决定系数R2从0.9711提升到0.9984,均方根误差从0.200%减小到0.0404 %。本文方法和研究结果可为铀元素含量的测量提供技术支撑。
Abstract
Objective

The nuclear industry is a strategic high-tech industry and an important cornerstone of national security. It involves various areas, such as ore exploration and mining, uranium extraction, isotope separation, reactor power generation, and spent fuel reprocessing. The uranium content of uranium ores is an important criterion for identifying uranium ore types and evaluating their developmental value. The rapid collection of uranium distribution information is necessary for geographical exploration. In particular, this is true for China, where uranium deposits are scattered and ore bodies are relatively small. Laser-induced breakdown spectroscopy (LIBS) is an atomic emission spectroscopy technique that involves irradiating the sample surface with nanosecond pulse lasers (typically at irradiance levels above GW/cm2). The irradiated material on the sample surface is rapidly heated, melted, vaporized, and partially ionized, forming laser-induced plasma (LIP). The elemental composition of the sample material can be measured by analyzing the emission spectra of the plasma. Fiber-optic LIBS (FO-LIBS) is an LIBS system that utilizes optical fibers for laser transmission and simultaneous collection of plasma emission spectra. It uses flexible, long optical fibers to transmit pulse lasers and spectral signals, which make it more suitable for complex and confined spaces in the field than conventional LIBS. Measurement distances can reach tens of meters. This study addresses the demand for the rapid, in-situ, and on-site detection of uranium in the nuclear industry and establishes a laboratory-based FO-LIBS system for investigating the evolution characteristics of uranium emission spectral lines in plasma under a helium atmosphere. Furthermore, it provides parameter optimization schemes and explores the matrix effects of uranium ore samples. A multivariate calibration method for quantitative analysis is proposed, which effectively improves calibration and prediction accuracy while ensuring model generalization performance. This provides a new approach for the rapid elemental analysis of ores.

Methods

We conducted experiments using natural samples and their mixtures to better align the results with practical applications. Spectra of the pressed samples were acquired using the FO-LIBS system. An air-blowing device was used to create a helium atmosphere, and the spectral information in a helium atmosphere was compared with that in an air atmosphere. The detection delay was optimized by comparing the signal-to-noise ratio, signal-to-background ratio, and net spectral intensity of the spectral lines. A multivariate linear calibration algorithm based on an internal standard method was proposed to address the matrix effects caused by the compositional differences among the samples. The model was fitted using partial least squares regression (PLSR) and a constrained genetic algorithm (GA), and the results were compared with calibration results based on spectral net intensity.

Results and Discussions

Among the U I 356.659 nm, U II 367.007 nm, and U II 409.013 nm lines in the uranium ore, only the U II 409.013 nm line exhibits a higher signal-to-noise ratio and is unaffected by interference from other lines at low mass fraction (Fig.3). In a helium atmosphere, the signal-to-noise ratio of U II 409.013 nm increases by 1.37 times from 13.29 to 31.45. Additionally, the signal-to-noise ratio reaches 8.9 at a mass fraction of 0.0726%. During the study of the variation in the detection delay using FO-LIBS in a helium atmosphere (Fig.4), the signal-to-noise ratio of the characteristic spectral lines remains above 10 until a delay of 1000 ns; however, it rapidly decreases to approximately 5 after a delay of more than 1000 ns. The signal-to-background ratio exhibits a peak of approximately 2.4 at a delay of 1000 ns and continues to increase subsequently when the delay is over 1200 ns, primarily owing to the rapid decay of the background intensity in the later stage of the plasma compared to those of the spectral lines. Therefore, a detection delay of 1000 ns is selected as the optimal value. Finally, a comparison of the results of the univariate calibration, multivariate linear regression using PLSR, and multivariate linear regression using constrained GA (Fig.8) shows that the prediction results obtained using multivariate linear regression are closer to the reference values than those obtained using univariate calibration based on spectral intensity alone. This indicates that the multivariate regression approach can correct for the matrix effects. The R2 (coefficient of determination) values of the calibration models based on PLSR and GA have both the training set and leave-one-out cross-validation (LOOCV) greater than 0.99, indicating the accuracy and robustness of these models. A comparison of PLSR and GA shows that the PLSR model exhibits superior calibration accuracy with a higher R2 and lower root mean square error in LOOCV. By constraining parameter k to positive values using the GA, the calibration accuracy decreases slightly; however, the relative standard deviation (RSD) decreases, resulting in improved prediction stability. The limits of detection and quantification are estimated as 142 mg/kg and 426 mg/kg, respectively.

Conclusions

This study investigates a uranium detection method based on FO-LIBS to meet the demand for rapid, on-site, and in-situ uranium detection in the nuclear industry. Among the dense spectra containing multiple elements, the uranium spectral line U II 409.013 nm is selected. The enhancement effect of the helium atmosphere on the uranium spectral line is explored. For the sample with a uranium mass fraction of 0.425%, the helium atmosphere improves the signal-to-noise ratio of the spectral line by 1.37 times. In addition, the detection delay of the system is optimized, and a peak in the signal-to-background ratio is observed at 1000 ns, which is determined to be the optimal delay for quantitative analysis. Under optimal conditions, the signal-to-noise ratio of the uranium spectral line is 8.9 in a sample with a mass fraction of 0.0726%. A multivariate linear regression model based on the internal standard method is proposed to address the matrix effect caused by differences in the chemical compositions of the natural samples in the experiments. The spectral lines of the matrix elements are introduced for calibration. The fitting parameters are obtained using PLSR and a constrained GA, with PLSR exhibiting superior quantitative performance in terms of R2 and RMSEC. The calibration model achieves an R2 of 0.9984 for uranium and an RMSEC of 0.0404%. Furthermore, the limit of detection for uranium using FO-LIBS is estimated to be 142 mg/kg, and the limit of quantification is 426 mg/kg.

1 引言

核工业是高科技战略产业,是****的重要基石1,涉及矿石勘探开采、铀提取、同位素分离、反应堆发电、乏燃料后处理等。铀元素在核材料中的测量是研究和生产过程中的重要共性技术,可为核燃料循环、核保障监督、核库存管理等提供重要支撑2。当前,铀元素的常规检测方法包括化学滴定法、电感耦合等离子体质谱法/光谱法(ICP-MS/OES)3-4等。这些方法虽然准确率高,但样品,特别是具有放射性的样品,处理流程比较复杂,处理大量样本时耗时较长。

激光诱导击穿光谱(LIBS)技术是一种原子发射光谱技术,其原理为:纳秒脉冲激光辐照样品表面(辐照度通常高于GW/cm2量级),样品被迅速加热后熔化、气化并被部分电离,形成激光诱导等离子体(LIP),通过分析等离子体发射光谱可以实现对样品中元素组分的定性识别和定量测量5。LIBS技术以其快速检测、无损、无需预处理等优势,在生物医学、冶金分析、环境检测等领域被广泛应用6-8

LIBS应用于铀元素测量时,由于等离子体中激发态铀粒子能量分布的能级比较广泛,自发衰减跃迁方式多样,因此其发射谱线的数量较多,而且强度低于铝、铜、铬等金属元素,很容易被其他谱线和背景噪声所覆盖9。近年来,在优化铀元素谱线的LIBS检测方面,研究人员对气氛条件10、激光辐照度11、二维光谱测量12、双脉冲系统13、空间约束14等开展了探索。2009年,Chinni等15在氩气气氛下研究了土壤基质中的铀谱线,他们通过优化气氛条件使检测限最终达到了0.26%。2012年,Kim等16在10 mJ的能量下,通过积累690幅光谱和使用铁光谱内标法,使铀元素的检测限达到了158 mg/kg。2016年,Skrodzki等17对比了双脉冲和单脉冲激光设置对铀谱线的增强效果,结果显示:双脉冲条件下U I 356.18 nm谱线的信噪比相比单脉冲条件下增加了1.5倍,等离子体的平均激发温度上升了10%。舒开强等18引入支持向量机和主成分回归模型校正铀矿基质效应造成的光谱抖动,将决定系数R2从0.93提升到0.99。张志威等19对铀元素含量不同的土壤样品和二氧化硅样品进行激发(结合光电双脉冲增强),使土壤中铀元素的检测限达到了108 mg/kg。清华大学的王哲课题组20先将光束整形至平顶模式,以有效降低等离子体屏蔽效应和波动性;然后在160 mJ激光能量下,采用LIBS技术对矿石中的铀元素进行探测,实现了21.2 mg/kg的检测限。

上述工作表明,铀元素检测限可达到百mg/kg等级21。但是,这些都是在实验室内采用高斯单模激光直接聚焦实现的,在实际应用中,含铀样品具有一定的放射性,高斯单模激光难以应用。光纤式激光诱导击穿光谱系统采用数十米长的柔性光纤传输脉冲激光和光谱信号,激光器和光谱可以远离待测样品,更适合应用于具有复杂排布的场景或狭小空间。英国Applied Photonics公司研制了核电用光纤式激光诱导击穿光谱(FO-LIBS)装置,其光纤长度达到了75 m,并成功完成了英国Hinkley Point B核电机组蒸汽过热器管道的Cu元素测量,检测限达到了1%,定量误差为15%22。为检查福岛核电站事故后核反应堆残骸的元素组分,日本开发了FO-LIBS装置,并采用该装置对水下反应堆模拟碎片样品进行了测量,结果显示Zr/Ce、Fe/Ce的检测限分别为0.0161、0.013923。在FO-LIBS系统中,光束经由多模光纤传输后呈“平顶”形貌,降低了聚焦光斑的峰值辐照度24,导致等离子体发光弱、痕量元素的探测灵敏度降低25。因此,采用FO-LIBS系统对铀元素进行检测更具挑战。

针对上述问题,本课题组建立了FO-LIBS系统,研究了环境气氛对铀发射谱线演化特性的影响规律,给出了优化的探测延时参数,探讨了铀矿样品的基质效应,提出了多元定标定量分析方法。所提多元定标定量分析方法在保证模型泛化性能的同时有效提高了定标和预测精度,为铀矿石的快速成分分析提供了一种新手段。

2 实验与方法

2.1 FO‑LIBS实验系统

所采用的实验系统示意图如图1所示,实验系统由激光器、光谱仪、时序控制器等硬件设备以及分束镜、透镜、光纤等光学器件组成。Nd∶YAG激光器(GKNPS-1064-15-10)发射的激光光束(波长为1064 nm,半峰全宽为8 ns,最大单脉冲能量为150 mJ)首先经过可变偏振分束器(VA5-1064)改变激光输出能量,然后由高透射比紫外熔融石英分束镜1、分束镜2(BSF10-C)分别采样,之后由光电二极管(DET10A, Thorlabs)、激光能量计(PE50BF-DIF-C)记录并控制时序以及调节并监测脉冲能量。透射光束通过透镜1 (LA4647,焦距f=20 mm)射入分束光纤的主光纤中(光纤长度为5 m,传输效率为83.4%,输出端激光能量控制为30 mJ),传输光束在接近样品的另一端射出,经过透镜2(LA4052,f=3.5 mm)后聚焦到样品表面激发等离子体。等离子体发光同步经透镜2成像于光纤馈出端面,然后经高紫外透射效率的信号光纤传输至以增强型电荷耦合器件(ICCD,DH334T)为探测器的中阶梯光谱仪(LTB,光谱范围为270~690 nm,分辨率约为12500),积分时间为1 μs。本实验系统的时序控制由数字延时发生器(DG645)实现。此外,本课题组为等离子体激发与演化提供了氦气氛围(通过向靶面吹扫实现),氦气流量由气体流量控制计控制,并被优化为5 L/min。

图 1. 铀元素检测实验系统

Fig. 1. Uranium element detection experimental system

下载图片 查看所有图片

为降低铀矿石样品中元素不均匀分布以及激光能量波动造成的影响,在每个位置累加20激光发次的光谱。本文报道的数据点均是对每种样品重复采集40组光谱后取平均值得到的。

2.2 样品制备

使用国家铀矿石标准物质系列样品开展实验,标准物质的铀含量在表1中给出。样品的制备方法:铀矿破碎、磨细、筛分后过200目筛,并在不锈钢转鼓内混匀2~3 h;将铀矿粉与聚乙烯醇黏合剂按4∶1的质量比进行混合(为激光烧蚀样品提供足够的表面强度),混合均匀后在30 MPa压力下压制3 min,完成样品制备。由于符合LIBS测量范围的天然样品较少且样品各组分含量相差过大,故而又将GBW04101、GBW04105这两种天然铀矿粉末混合,制作了5类定标样品。样品编号及含量信息如表2所示。

表 1. 标准样品中铀元素及其他组分的质量分数

Table 1. Mass fraction of uranium element and other components in standard samples

SampleMass fraction /%
USiO2Al2O2CaOMgO
GBW041013.2981.316.290.8060.312
GBW041030.21950.4213.3913.750.561
GBW041040.07265.011.5027.9218.74
GBW041050.42556.8417.004.861.18

查看所有表

表 2. 混合样品的混合比例和铀含量

Table 2. Mixing ratio and uranium content of mixed samples

NumberMass fraction of uranium /%Mass fraction of GBW04101 /%Mass fraction of GBW04105 /%
12.619779.1720.83
22.016460.4239.58
31.681350.0050.00
41.346239.5860.42
50.742920.8379.17

查看所有表

3 结果与讨论

3.1 系统参数优化

由于铀元素谱线强度低且数量多,而且矿石样本中的元素种类丰富,进一步增加了光谱识别的难度,因此,铀谱线易与其他谱线重叠或被背景遮盖。为了确定高强度的铀元素谱线,选择铀质量分数为3.29%的矿石样本进行分析,其原始光谱如图2所示。经过与NIST数据库比对,筛选出U I 356.659 nm、U II 367.007 nm、U II 409.013 nm三条谱线用于本文的量化研究。

图 2. GBW04101样本中铀及其他元素谱线的识别

Fig. 2. Identification of spectral lines of uranium and other elements in GBW04101 sample

下载图片 查看所有图片

相比于单模高斯光束直接聚焦在样品表面的常规LIBS,FO-LIBS中光纤馈出激光时的发散角导致了更低的能量密度,需要采取额外的增强手段。首先通过调节环境气氛和系统参数的方法提高谱线的信噪比。图3展示了铀质量分数分别为0.0726%、0.219%、0.425%的样本在空气和氦气氛围下的三条铀谱线。氦的电离能(24.6 eV)大于氮(15.6 eV)和氧(12.1 eV)的电离能,故而氦气气氛将会减轻气体击穿,从而减小电子数密度,降低整谱的连续发射。同时,氦气气氛也会减弱等离子体屏蔽效应,使更大能量的激光光束辐照表面,提升激光-靶的耦合效率,增大烧蚀采样率26。由图3(c)、(d)可以看到氦气环境中的连续辐射相比空气环境中的有所降低。由图3(e)、(f)可以看出,相比空气环境,氦气环境下的光谱既出现了背景强度的降低,也出现了铀特征谱线的增强。综上所述,尽管三条谱线在氦气环境下的净强度都有所增加,但只有U II 409.013 nm谱线在铀含量较低时不受其他谱线的干扰,而且保持了较高的信噪比。

图 3. 铀质量分数分别为0.0726%、0.219%、0.425%的样本在空气和氦气气氛下的铀谱线。(a)(b)U I 356.659 nm;(c)(d)U II 367.007 nm;(e)(f)U II 409.013 nm

Fig. 3. Uranium spectral lines of the samples with mass fraction of 0.0726%, 0.219%, and 0.425% in air and He atmosphere. (a)(b) U I 356.659 nm; (c)(d) U II 367.007 nm; (e)(f) U II 409.013 nm

下载图片 查看所有图片

在氦气环境中,铀质量分数为0.425%的样本的U II 409.013 nm谱线的信噪比为31.45,相比空气中的信噪比(13.29)增大了1.37倍;同时,铀质量分数为0.0726%的样本的U II 409.013 nm谱线的信噪比达到了8.9,满足谱线信噪比大于3的定性识别要求。

图4展示了氦气气氛中铀质量分数为0.425%的样本中的铀谱线及其信背比、信噪比、净强度随探测延时的变化,其中背景强度采用U II 409.013 nm谱线两侧408.817~408.98 nm和409.114~409.15 nm范围内光谱强度的平均值,噪声采用背景区间强度的标准差。可以观察到:随着延时增加,等离子体逐渐冷却,光谱整体强度和铀谱线净强度持续下降;在延时达到1000 ns之前,特征谱线的信噪比(SNR)维持在10以上,但延时超过1000 ns后,信噪比快速下降至5左右;信背比(SBR)在延时达到1000 ns时出现约2.4的峰值,并在延时超过1200 ns后继续增大(这主要是由于在等离子体后期背景强度的减弱速度高于谱线的减弱速度)。综合考虑上述因素后,选择1000 ns作为最佳的探测延时。

图 4. 氦气气氛中铀谱线及其信背比、信噪比和净强度随探测延时的变化。(a)谱线随探测延时的变化;(b)谱线信背比、信噪比和净强度随探测延时的变化

Fig. 4. Variations of uranium spectral lines and their signal-to-background ratio (SBR), signal-to-noise ratio (SNR), and net intensity with detection delay in He atmosphere. (a) Variation of spectral line with detection delay; (b) variation of SBR, SNR, and net intensity of the spectral lines with detection delay

下载图片 查看所有图片

3.2 基质校正算法及铀元素定标

不同类型矿石在成分、微观结构上的差异使得其在同等制样条件下的表面强度存在显著差异,导致明显的基质效应,进而导致光谱信号发生波动。图5展示了三种天然铀矿样本的光谱,可以看出:GBW04101样本的光谱轮廓与其他样本差异较大;尽管GBW04103和GBW04105样本有相似的光谱轮廓,但它们的整体强度有显著差异。基质效应对LIBS定标模型的精确度有很大影响,为了降低基质效应,研究人员通常在具有相同基质的标准样品中添加不同量的目标元素,或者使用内标法来定标痕量元素谱线相对于大量元素谱线的变化。然而,传统的内标法在预测基质差异较大的样本时往往效果不佳,因为很难找到一种大量或恒量的元素作为内部标准。本文提出了一种基于内标法原理的多元线性回归模型,通过引入多个基质元素来提升定标模型对基质差异的校正能力。

图 5. 不同天然样本的光谱比较

Fig. 5. Spectra comparison of different natural samples

下载图片 查看所有图片

根据激光诱导等离子体理论,原子特征谱线发射强度为

Iji=hc4πλjiNsUj(Te)gjAjiexp-EjKBTe

式中:Iji为谱线发射强度;λji为发射谱线的波长;Aji为跃迁概率;Ej为激发态能级的能量;gj为激发态的统计权重;Te为等离子体温度;KBh、c分别为玻尔兹曼常数、普朗克常数和光速;Ns为等离子体中该元素的原子或离子的数密度;UjTe)为相应的跃迁粒子的配分函数。上述参数均可从美国国家标准与技术研究院(NIST)原子光谱数据库中查询得到。

通常采用内标法来减小测量过程中的不稳定性。内标法表达式为

I1I2=kn1n2

式中:I为谱线强度;n为元素的质量分数;k为定标模型的斜率。由于天然矿石样本中的每种元素的含量有明显的区别,因此,基于内标法原理,在引入除目标元素外的多个基质元素后,将式(2)变形,得到

k1I1IU+k2I2IU+=n1+n2+nU

式中:IUIi分别为目标元素和基质元素的光谱强度;nUni分别为目标元素和基质元素的质量分数。当选取的基质元素质量分数之和接近100%时,式(3)等号右侧分式的分子近似为1-nU

在诸如式(3)形式的多元定标模型中,参数群{k}通常需要采用式(2)所示的内标法分别计算得出,计算时不仅需要已知各个基质元素的含量,还要求所使用的基质元素都与目标元素有至少一条良好满足式(2)的内标谱线,这在实际应用中很难实现。因此,笔者通过多元线性拟合来获得式(3)中的参数群{k}。

此外,定标所用样本数量有限,若同时使用数量过多的谱线进行拟合,则可能会造成严重的过拟合现象。因此,应适当限制同时使用的基质元素的数量,并采用留一法交叉验证(LOOCV),以便在不影响定标精度的前提下保证模型的泛化能力。综合考虑后,笔者选择了3个元素。同时,笔者建立了一套通过查找谱线库来训练定标模型的流程,如图6所示。

图 6. 基质校正训练流程图(R2:决定系数;RMSECV:交叉验证的均方根误差;ε1:准许的最小决定系数;ε2:准许的最大均方根误差)

Fig. 6. Flow chart of matrix correction training (R2: coefficient of determination; RMSECV: root mean square error of cross verification; ε1: acceptable minimum coefficient of determination; ε2: acceptable maximum root mean square error of cross verification)

下载图片 查看所有图片

在解决多元线性拟合问题时,笔者对比了偏最小二乘法回归(PLSR)和遗传算法(GA)的拟合效果。其中:PLSR通过构建自变量与因变量相互正交的特征向量来建立自变量与因变量的特征向量之间的一元线性回归关系,不仅克服了共线性问题,还在选取特征向量时强调了自变量对因变量的解释和预测作用,去除了对回归无益的噪声的影响,使模型包含最少的变量数。GA是在自然进化理论启发下发展的一系列搜索算法,它通过模仿自然选择和繁殖的过程,为涉及搜索、优化和学习的各种问题提供高质量的解决方案;由于类似于自然进化,GA可以克服传统搜索和优化算法遇到的一些障碍,在解决具有大量参数和复杂数学表示形式的问题上具有巨大优势。对于基于式(3)的训练模型,GA可以对k值的范围进行约束,使其相对于无约束的PLSR在迭代过程中更容易接近方程的真实解。

接下来分别采用PLSR、GA获取拟合参数。采用PLSR得到的最优谱线组合为Si I 298.7645 nm、Fe I 370.9246 nm和Ca I 422.673 nm,这些谱线的信息如表3所示。GA拟合参数时全局收敛速度慢且难以得到全局最优解27,因此,选择PLSR优化的谱线组合进行GA拟合。

表 3. 输入多元线性回归模型的元素谱线信息

Table 3. Information of element spectral lines used in multiple linear regression model

ElementWavelength /nmLower energy level transitionUpper energy level transitionUpper energy level /cm-1
Si I298.76453s23p23s23p4s39760.285
Fe I370.92463d24F)4s3d24F)4p34328.752
Ca I422.6733p64s23p64s4p23652.304

查看所有表

基于多元线性回归模型和表1中的标准样品,对铀元素建立定标模型。为了降低背景造成的额外影响,使用扣除背景后的铀元素谱线,如图7所示。采用LOOCV来评估定标模型的泛化能力:在一系列样本中取出一个样本,用其他样本作为训练集建立定标模型,对取出的样本进行预测;循环进行上述操作,直至所有样本都被取出。

图 7. 铀质量分数不同的标准样品和混合样品的U II 409.013 nm谱线

Fig. 7. U II 409.013 nm spectral line of standard and mixed samples with different uranium mass fractions

下载图片 查看所有图片

模型的评价指标包括决定系数R2、校准均方根误差(在公式中记为eRMSEC)、交叉验证的均方根误差(在公式中记为eRMSECV)、相对标准偏差(在公式中记为dRSD)、检测限(LOD,在公式中记为fLOD)、定量限(LOQ,在公式中记为fLOQ)和,它们的计算公式分别为

R2=i=1n(y̑i-y¯)2i=1n(yi-y¯)2eRMSEC=i=1m(y̑i-yi)2meRMSECV=i=1Ncv(y̑i-yi)2NcvdRSD=j=1n(y̑j-y¯)2(n-1)y¯fLOD=3σbfLOQ=3fLOD

式中:n为每m次检测的质量分数的数量;Ncv为交叉验证的样本数;y̑iyi分别表示第i个质量分数下的预测值和实际值;y̑j表示第j次测量下的预测值;y¯为上述结果的平均值;σ为背景噪声的标准差;b为定标曲线的斜率。图8展示了单变量定标、PLSR拟合的多元线性回归、GA拟合的多元线性回归的定标曲线,具体信息在表4中给出。

图 8. 定标曲线。(a)传统光谱信息定标;(b)PLSR基质校正;(c)GA回归基质校正

Fig. 8. Calibration curves. (a) Traditional spectral information calibration; (b) PLSR matrix correction; (c) GA regression matrix correction

下载图片 查看所有图片

表 4. 三种回归方法下的定标指标

Table 4. Calibration indices under three regression methods

MethodR2RMSEC /%RMSECV /%RSD /%
CalibrationLOOCV
Spectral information calibration0.97110.96850.2000.40118.70
PLSR0.99840.99830.04040.044122.24
GA0.99070.99190.08950.095319.65

查看所有表

相对于光谱净强度单变量定标,多元线性回归的预测结果更接近参考值,这意味着其能够达到校正基质效应的目的。表4中,通过PLSR和GA得到定标模型的训练集和LOOCV下的定标决定系数R2均大于0.99,可以得到精确且泛化能力强的定标模型。对比PLSR和GA两种回归算法可知,PLSR拟合模型的定标精度更高,在LOOCV中具有更高的R2和更低的均方根误差。通过GA约束参数群{k}为正值,定标精度略微降低,但RSD减小,具有更好的预测稳定性。单变量定标得出的模型为y=2638.1x,背景信号的标准差为12.49,根据式(8)、(9)计算可得检测限和定量限分别为142 mg/kg和426 mg/kg。

本研究与其他研究的定标结果比较如表5所示。可以看出,提高激光能量有利于降低检测限。本文基于光纤式LIBS系统,通过发展多元定标算法,在10 mJ级激光能量下实现了与实验室相近的结果。本研究工作在铀矿石测量以及辐照、水下等严苛环境中核电站反应堆异物测量、腐蚀产物测量等方面都具有广阔的应用前景。

表 5. 本研究与其他研究的实验参数和定标结果比较

Table 5. Comparison of experimental parameters and calibration results between this study and other studies

YearLIBS setupSpectral lineSampleLOD /(mg·kg-1
200915Focus directlyArU II 409.013 nmSoil2600
201128Focus directly10 mJ,airU I 358.49 nmSiO2150
201216Focus directly,10 mJ,airU I 356.66 nmOre158
20169Focus directly,10 mJ,CO2U II 409.013 nmSiO2272
202121Focus directly,11.5 mJ,airU II 304.41 nmSoil38
202220Focus directly,160 mJ,airU II 409.013 nmOre21
OursFocus after 5 m fiber,30 mJ,HeU II 409.013 nmOre142

查看所有表

4 结论

面向核工业对铀元素快速、原位、现场检测的需求,笔者研究了基于FO-LIBS的铀元素含量检测方法。在具有多种元素的密集光谱中筛选出U II 409.013 nm铀谱线,探究了氦气气氛对铀谱线的增强效果。在铀元素质量分数为0.425%的矿石样本中,氦气气氛可使谱线的信噪比提高1.37倍。同时,优化了系统探测延时,在1000 ns处信背比出现峰值,故将1000 ns设为定量分析的最佳延时。在最优条件下,铀质量分数为0.0726%的矿石样本中铀谱线的信噪比达到了8.9。

为解决实验中天然样品化学成分差距导致的基质效应,提出了基于内标法的多元线性回归模型,引入基质元素谱线加以校正。分别通过PLSR和有约束的GA获取拟合参数,结果显示,PLSR在R2和RMSEC等定量指标上表现更加优异:对铀元素的定标决定系数R2达到了0.9984,RMSEC为0.0404%。另外,计算得到FO-LIBS对铀的检测限为142 mg/kg,定量限为426 mg/kg。本文在10 mJ级激光能量下达到了与实验室直接聚焦相近的检测限,对铀矿石、反应堆异物和核电站事故堆芯产物中铀元素的现场原位检测提供了技术支撑。

参考文献

[1] 潘建均, 康椰熙, 李筱珍, 等. 我国核工业标准化研究的演进、热点与前沿: 基于中国知网1980—2021年文献的科学计量分析[J]. 科技管理研究, 2022, 42(9): 31-40.

    Pan J J, Kang Y X, Li X Z, et al. Evolution, hotspots, and frontiers of standardization research in China’s nuclear industry: scientific econometric analysis based on the literature of CNKI from 1980 to 2021[J]. Science and Technology Management Research, 2022, 42(9): 31-40.

[2] Wu J, Qiu Y, Li X W, et al. Progress of laser-induced breakdown spectroscopy in nuclear industry applications[J]. Journal of Physics D: Applied Physics, 2020, 53(2): 023001.

[3] 付琳, 王玉学, 魏代东, 等. 电感耦合等离子体光谱-质谱法联合测定铀化合物粉末中20种杂质元素[J]. 化学分析计量, 2022, 31(8): 17-21.

    Fu L, Wang Y X, Wei D D, et al. Determination of 20 impurity elements in uranium compound by inductively coupled plasma spectroscopy and mass spectrometry[J]. Chemical Analysis and Meterage, 2022, 31(8): 17-21.

[4] 侯列奇, 王树安, 李洁, 等. 电感耦合等离子体原子发射光谱法测定铀-钼合金中15种微量杂质元素[J]. 理化检验-化学分册, 2007, 43(3): 179-181, 184.

    Hou L Q, Wang S A, Li J, et al. ICP-AES determination of micro-amounts of 15 elements present as impurities in uranium-molybdenum alloy[J]. Physical Testing and Chemical Analysis Part B (Chemical Analysis), 2007, 43(3): 179-181, 184.

[5] 杨雪, 张丹, 陈安民, 等. 聚焦透镜到样品表面的距离对激光诱导硅等离子体原子谱线强度和离子谱线强度的影响[J]. 中国激光, 2019, 46(11): 1111001.

    Yang X, Zhang D, Chen A M, et al. Influence of distance between focusing lens and sample surface on atomic line and ionic line intensities of laser-induced silicon plasmas[J]. Chinese Journal of Lasers, 2019, 46(11): 1111001.

[6] 李铸, 张庆永, 孔令华, 等. 基于激光诱导击穿光谱与随机森林识别GCr15钢的硬度[J]. 中国激光, 2022, 49(9): 0911002.

    Li Z, Zhang Q Y, Kong L H, et al. Hardness characterization of GCr15 steel based on laser-induced breakdown spectroscopy and random forest[J]. Chinese Journal of Lasers, 2022, 49(9): 0911002.

[7] 孟德硕, 赵南京, 刘文清, 等. 激光诱导击穿光谱结合标准加入法定量检测土壤中Cr[J]. 中国激光, 2014, 41(7): 0715002.

    Meng D S, Zhao N J, Liu W Q, et al. Quantitative measurement of Cr in soil using laser induced breakdown spectroscopy combined with standard addition method[J]. Chinese Journal of Lasers, 2014, 41(7): 0715002.

[8] 陈兴龙, 董凤忠, 陶国强, 等. 激光诱导击穿光谱在地质录井岩性快速识别中的应用[J]. 中国激光, 2013, 40(12): 1215001.

    Chen X L, Dong F Z, Tao G Q, et al. Fast lithology identification by laser-induced breakdown spectroscopy[J]. Chinese Journal of Lasers, 2013, 40(12): 1215001.

[9] Barefield J E, Judge E J, Campbell K R, et al. Analysis of geological materials containing uranium using laser-induced breakdown spectroscopy[J]. Spectrochimica Acta Part B: Atomic Spectroscopy, 2016, 120: 1-8.

[10] Taylor N R, Phillips M C. Differential laser absorption spectroscopy of uranium in an atmospheric pressure laser-induced plasma[J]. Optics Letters, 2014, 39(3): 594-597.

[11] Sarkar A, Alamelu D, Aggarwal S K. Laser-induced breakdown spectroscopy for determination of uranium in thorium-uranium mixed oxide fuel materials[J]. Talanta, 2009, 78(3): 800-804.

[12] Phillips M C, Brumfield B E, LaHaye N, et al. Two-dimensional fluorescence spectroscopy of uranium isotopes in femtosecond laser ablation plumes[J]. Scientific Reports, 2017, 7: 3784.

[13] Zhang Z, Wu J, Hang Y H, et al. Quantitative analysis of chlorine in cement pastes based on collinear dual-pulse laser-induced breakdown spectroscopy[J]. Spectrochimica Acta Part B: Atomic Spectroscopy, 2022, 191: 106392.

[14] 李佶龙, 王傲松, 刘彦章, 等. 平板约束下激光诱导等离子体的动态仿真[J]. 光子学报, 2020, 49(2): 0230002.

    Li J L, Wang A S, Liu Y Z, et al. Dynamic simulation of laser-produced plasma under plate constraint[J]. Acta Photonica Sinica, 2020, 49(2): 0230002.

[15] Chinni R C, Cremers D A, Radziemski L J, et al. Detection of uranium using laser-induced breakdown spectroscopy[J]. Applied Spectroscopy, 2009, 63(11): 1238-1250.

[16] Kim Y S, Han B Y, Shin H S, et al. Determination of uranium concentration in an ore sample using laser-induced breakdown spectroscopy[J]. Spectrochimica Acta Part B: Atomic Spectroscopy, 2012, 74/75: 190-193.

[17] Skrodzki P J, Becker J R, Diwakar P K, et al. A comparative study of single-pulse and double-pulse laser-induced breakdown spectroscopy with uranium-containing samples[J]. Applied Spectroscopy, 2016, 70(3): 467-473.

[18] 舒开强, 陈友元, 彭郑英, 等. 铀矿中多目标元素的激光诱导击穿光谱定量分析方法研究[J]. 分析化学, 2023, 51(7): 1195-1207.

    Shu K Q, Chen Y Y, Peng Z Y, et al. Laser-induced breakdown spectroscopy for quantitative analysis of multi-target elements in uranium ore[J]. Chinese Journal of Analytical Chemistry, 2023, 51(7): 1195-1207.

[19] 张志威, 邱荣, 姚胤旭, 等. 铀元素的激光诱导击穿光谱测量分析[J]. 光谱学与光谱分析, 2023, 43(1): 57-61.

    Zhang Z W, Qiu R, Yao Y X, et al. Measurement and analysis of laser-induced breakdown spectrum of uranium element[J]. Spectroscopy and Spectral Analysis, 2023, 43(1): 57-61.

[20] Ji J X, Song W R, Hou Z Y, et al. Raw signal improvement using beam shaping plasma modulation for uranium detection in ore using laser-induced breakdown spectroscopy[J]. Analytica Chimica Acta, 2022, 1235: 340551.

[21] Han S K, Park S H, Ahn S K. Quantitative analysis of uranium in electro-recovery salt of pyroprocessing using laser-induced breakdown spectroscopy[J]. Plasma Science and Technology, 2021, 23(5): 055502.

[22] WhitehouseA I, YoungJ, EvansC, et al. Remote compositional analysis of spent-fuel residues using laser-induced breakdown spectroscopy[EB/OL]. [2023-03-05]. https:∥www.osti.gov/biblio/826347.

[23] Saeki M, Iwanade A, Ito C, et al. Development of a fiber-coupled laser-induced breakdown spectroscopy instrument for analysis of underwater debris in a nuclear reactor core[J]. Journal of Nuclear Science and Technology, 2014, 51(7/8): 930-938.

[24] 吕启深, 邱岩, 唐峰, 等. 采用光纤传输激光的激光诱导击穿光谱系统参数[J]. 高电压技术, 2020, 46(9): 3301-3310.

    Lü Q S, Qiu Y, Tang F, et al. Parameters of laser-induced breakdown spectroscopy system using fiber-optic transmission laser[J]. High Voltage Engineering, 2020, 46(9): 3301-3310.

[25] Qiu Y, Wu J, Zhang Z, et al. Comparisons of laser-produced plasma in atmosphere between fiber-delivery and direct-focusing laser pulse[J]. Spectrochimica Acta Part B: Atomic Spectroscopy, 2019, 155: 12-23.

[26] Iida Y. Effects of atmosphere on laser vaporization and excitation processes of solid samples[J]. Spectrochimica Acta Part B: Atomic Spectroscopy, 1990, 45(12): 1353-1367.

[27] 曹椋焱, 李光布, 李景辉. 遗传算法的分析及其改进[J]. 计算机仿真, 2009, 26(7): 228-231.

    Cao L Y, Li G B, Li J H. Analysis of genetic algorithm and its modification[J]. Computer Simulation, 2009, 26(7): 228-231.

[28] Jung E C, Lee D H, Yun J I, et al. Quantitative determination of uranium and europium in glass matrix by laser-induced breakdown spectroscopy[J]. Spectrochimica Acta Part B: Atomic Spectroscopy, 2011, 66(9/10): 761-764.

郭歆宇, 吴坚, 时铭鑫, 朱欣欣, 邱岩, 李京徽, 吴迪, 宋自远. 基于多元定标法的矿物铀元素光纤式激光诱导击穿光谱定量分析[J]. 中国激光, 2024, 51(8): 0811001. Xinyu Guo, Jian Wu, Mingxin Shi, Xinxin Zhu, Yan Qiu, Jinghui Li, Di Wu, Ziyuan Song. Quantitative Measurement of Uranium in Ore Using Fiber‑Optic Laser‑Induced Breakdown Spectroscopy and Multivariate Calibration[J]. Chinese Journal of Lasers, 2024, 51(8): 0811001.

引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!